Career Guidance - Prod Mgmt vs Data Science

Amazon mapz
Jul 3 11 Comments

Hello folks,

I am in a mid-life career crisis here so I was hoping to receive some guidance from fellow Bliind'ers who know better.

My current official job title is Product Manager and it has been so for the past 3 years with the company. However I have rarely got to do any real work related to Prod Mgmt. My actual job responsibilities have been intermittently switching between that of a Business/Data Analyst/Business Intelligence Engineer which means my team has been dependent on me mostly for my skillset on SQL and reporting tools with some adhoc program management.

Now I am at a point where I do not want to continue doing so and instead focus on my long term career growth which I believe my current role does not offer. Based on my experience with the respective teams, I feel Prod Mgr and Data Scientist are two different roles that offer a lot of challenge and responsibility for individual growth, and in the long term will help me develop into a Director/VP role. Although both roles require completely different skillset and parameters to be successful, I am trying to gauge each of them and hoping if I could find a middle ground where I can do a mix of both.

I did my Masters in MIS so I have sufficient educational background and experience to pursue either stream. I am interested in Prod Mgmt because I like the idea of owning and defining a product, preparing the roadmap and shaping it to completion by partnering with the Tech teams. I like the satisfaction of a being an independent 'owner' and the responsibility associated with it. However, my only fear is the amount of 'creativity' and 'document writing' that is desired in such roles. I constantly come across interviews from FB, Google where Prod Mgrs are expected to come up with their own suggestions for a product and define the end to end design. I am not sure where I can develop such expertise and remain successful when competing with MBA holders who are better at such stuff including writing top quality documents! Also, I am on a H-1B visa, so with the flurry of specialty occupation RFE exceptions raised recently, I am not sure if I should take the risk of sticking to a purely prod mgmt role when I cannot demonstrate any technical skills.

When it comes to Data science, I feel slightly more confident about my skillset because of my current work experience and prior work experience as a developer in Python. However, I realize there is a steep learning curve to this as well as recently I have noticed that HM's expect the candidates to know more about ML algorithms and statistics in addition to basic coding skills. Although I feel its relatively easy to learn these skills (compared to 'creative thinking' and 'doc writing'), my concern is if these skills will get outdated too soon! Also I feel that Data Science is a vast and an endless domain. With the evolution of Deep learning & Reinforcement learning, I feel that an individual can learn & practice for years and still not be a Master of data science, or would be lacking in one or the other skills. For example, if one team expects a Data Scientist to know Machine learning, the same cannot be true for a different team who may expect the candidate to know more about Deep learning/statistical analysis. Also, I feel the current demand for Machine & Deep Learning skills is just a phase and the tools (Python language, SQL, scikit-learn) will get outdated soon enough and in the end we will need to re-learn in a new set of tools/skills over the next decade. Additionally, I am not aware what does a 10/15 year roadmap looks like for a Data Scientist. Do they eventually grow to a Principal Data Scientist role? I haven't come across any person who is a VP of Data science.

I realize that the views I have shared above could be entirely misguided and inaccurate but I will be glad to hear everybody's thoughts on this so that I can get a better picture. Thanks!

comments

Want to comment? LOG IN or SIGN UP
TOP 11 Comments
  • Microsoft UMbR31
    How old are you and how many yoe you got?
    Jul 3 3
    • Amazon mapz
      OP
      31, 7 yoe, 4 as a SDET
      Jul 3
    • Microsoft UMbR31
      I'd get a product manager job in another Amazon team, where you are playing the product role. If there's no rush, you probably try interviewing for the next 6 months? Don't worry about H1B RFEs, if you stay with megacorps, they'll find a way.

      On why I don't recommend data science:
      1. I feel like two kinds of people might be needed a few years from now: A: ability to create ML infra, strong in math, ship useful APIs, creating tools and frameworks, work at giant corp B: business analyst users who drag and drop / SQL queries, using the tools made by giant co etc. I think this will become like databases on the cloud or data visualization tools like powerbi (which is obviating the need for people with these skills). If you're likely to be type B, life gonna be hard.

      2. Every change in perceived 'type of job' sets you back by some years.

      3. Since you're 7yoe on H1B, I'm assuming you're Indian. If, for whatever reason, you have to relocate, product manager at American corp is a more lucrative credential than Data scientist.
      Jul 3
    • Quora iSjP52
      Really agree with the above analysis. If you think of yourself as a generalist, go to PM. Or maybe DS management
      Jul 7
  • Amazon / Product GreyBar
    Data Science is a relatively new field as compared to Product Mgmt and hence you don’t see VPs of Data Science all around but I will not be surprised if VP of Data Science and Chief Data Officers become commonplace in another 10 years or so. Several companies have started hiring dedicated Head of Analytics and Data Science and as organisations realise the value of these roles, these roles will grow in stature.
    That being said, if your long term career path is to get to a senior General Management position, which most Amazon VPs are, you need to gradually move towards a role that gives you a higher degree of responsibility in terms of owning business goals, which the PM track is more likely to offer.
    Jul 3 0
  • American Express vista133
    Your view is structured and contains a lot of assumptions. Keep the structure, learn some strategy principles and question your assumptions... you are already a good PM.
    Jul 3 0
  • Intel / Product starlens
    You need to not let others define your role for you -- assert yourself into providing PM value to your team beyond being their resident data scientist.

    I once was in a situation where my eng team realized I had strong technical skills and wanted me to spend most of my time on eng work & program management (I'm really organized), I had to push back.

    As a result of my persuasion, the team realized they needed to hire people to perform the extra work I was covering for, and I was subsequently able to focus on actually defining product strategy.
    Jul 3 3
    • Microsoft UMbR31
      This sounds like an interview answer. Wonder what happened in reality...
      Jul 3
    • Intel / Product starlens
      You are correct that it sounds like an interview answer. Don't worry, I too noticed it immediately after typing the last paragraph but was too lazy to correct it.

      What happened in real life was I created a spreadsheet where I listed the product strategy work that wasn't being done and identified how this was compromising our ability to build a good product. I shared the spreadsheet with my eng lead and we went through each item during a 1:1.

      Generally I spend a lot of time building a good relationship with my eng lead(s) so that I can elicit empathy out of him/her when I need to. I also apply a purely logical approach sometimes -- I strategically push back against being assigned low priority action items (by listing other high priority work that would go undone). I believe that a PM should be pleasant to work with but should not be a walkover.
      Jul 4
    • New / Product
      WcBB32

      New Product

      PRE
      BNP Paribas
      WcBB32more
      “I believe that a PM should be pleasant to work with but should not be a walkover.”

      100%. However it is a very difficult balance to strike. (Source: 3-yoe PM with non-technical background).

      In addition to logical approach, I also find that open communication helps a lot with building trust with eng leads.
      Jul 16
  • New / Product viajero
    Go with Data Science for sure there is much more future opportunities for you and much greater earning potential and more chances to work remotely. I’m a PM and it gets hectic and tiring getting info from other people and trying to get a product out.
    Jul 3 0